In the current project, the techniques of classic epidemiology and of genetic epidemiology are being used to identify determinants of type 2 diabetes, its risk factors and its complications. The effects of genetic factors are assessed in other projects (DK069028-24, Genetic Epidemiology of Diabetes and Obesity and DK069094-06, Genetic Epidemiology of Diabetic Complications); the present study focuses on molecular processes that are downstream of genetic factors, such as gene transcription, protein expression and cellular metabolism. Longitudinal studies are being conducted in individuals at high risk for type 2 diabetes in urban Phoenix. Samples are collected for studies of genomic DNA, and these samples are also suitable for studies of epigenetic factors, such as DNA methylation and histone modification. Samples are collected that suitable for RNA extraction and for large scale studies of protein levels and metabolomics. Information on family membership is collected for inclusion in family-based analyses. Recent analyses of genome-wide transcription in skeletal muscle tissue have shown that a small number of transcripts have a clearly bimodal frequency distribution; the vast majority of these are accounted for by the effect of a genetic polymorphism in close proximity the region encoding for the transcript. Analyses using empirical identity by descent estimates from dense genotypic data suggest that a large amount of the heritability in gene expression is due to the action of variants within or near the region coding for the transcipt. Large scale genome-wide transcription profiles are also being measured in peripheral blood samples in collaboration with investigators from Texas Biomedical Research Institute. Initial analyses have shown a large number of transcripts associated with age, sex and markers of inflammation (C-reactive protein), and these data are being compared with those from other cohorts in a collaborative meta-analysis. . Recent studies have shown the antibodies to prefoldin2 (PFLD2) may predict risk of young-onset diabetes and give suggestive evidence that many persistent organic pollutants may also predict diabetes risk; additional replication studies are currently being planned. Analyses of the relationship of transcription in peripheral blood samples with diabetes, obesity and related factors are currently in process, as is the assessment of heritability and parent-of-origin effects. Recruitment of additional individuals and families and longitudinal follow-up are also ongoing.